Social media use and cosmetic surgery consideration among Chinese university students: a cross-sectional test of serial mediation through beauty ideal internalization and self-objectification

1 Introduction

Cosmetic surgery consideration has become an increasingly salient social and public health concern. Across diverse cultural contexts, the growing normalization of aesthetic procedures reflects heightened appearance pressures and intensified body dissatisfaction. Beyond individual decision-making, cosmetic surgery consideration signals broader psychological and sociocultural dynamics that shape how individuals evaluate and modify their bodies. Despite its rising prevalence and potential mental health implications, the underlying psychological mechanisms driving cosmetic surgery consideration remain insufficiently clarified. One factor associated with cosmetic surgery consideration is social media use. Previous research has examined specific social-media practices and appearance-related content exposure, including Highly Visual Social Media, image-based activities, and photo-editing behavior (Di Gesto et al., 2022; Hermans et al., 2022; Mironica et al., 2024). These constructs are distinct from the general self-reported social media use intensity measured in the present study. Accordingly, the present study does not assess platform-specific mechanisms, algorithmic curation, or exposure to appearance-focused content. However, the influence of social media is unlikely to operate through a simple direct effect. Communication research conceptualizes social media use as individuals’ routine engagement with digital platforms, encompassing usage frequency, intensity, and interactive participation (Ellison et al., 2007; Kross et al., 2013). Within such environments, repeated exposure to idealized and filtered representations may facilitate beauty ideal internalization—that is, the adoption of socially constructed appearance standards as personal evaluative criteria (Keery et al., 2004; van den Berg et al., 2002; Wang et al., 2022). Objectification theory further posits that internalized beauty norms may shift individuals toward adopting an observer's perspective on their own bodies, resulting in self-objectification characterized by persistent appearance monitoring and body shame (Fredrickson and Roberts, 1997; Moradi and Huang, 2008). These processes may cumulatively heighten cosmetic surgery consideration. At the same time, contemporary social media environments may contain both appearance-promoting and corrective content. Alongside idealized imagery, platforms also disseminate corrective narratives such as body-positivity movements, authenticity campaigns, and surgical risk disclosures. Given the coexistence of idealized and corrective discourses on social media, its relationship with cosmetic surgery consideration may reflect competing psychological pathways rather than a purely linear process. Exploring these dual mechanisms is therefore essential for understanding the complex relationship between social media use and cosmetic surgery consideration.

Objectification Theory, proposed by Fredrickson and Roberts (1997), constitutes a pivotal theoretical lens for explaining individuals’ body cognition shifts and appearance-related behaviors. This theory posits that in visually dominated socio-cultural contexts, persistent exposure to media messages framing bodies as “ an object to be looked at” leads individuals to gradually adopt and internalize external aesthetic standards. Consequently, they come to perceive their own bodies as objects for others’ evaluation—a process termed self-objectification. This process subsequently triggers outcomes such as body surveillance and body shame, ultimately influencing appearance-related attitudes and behaviors. The theoretical framework emphasizes significant logical linkages among three elements: external environment, individual psychological mechanisms, and appearance-related behaviors. Within this study, self-objectification and beauty ideal internalization serve as the psychological mediating mechanisms of this process. In addition to objectification theory, upward social comparison theory (Festinger, 1954) provides an important perspective for understanding how exposure to appearance-related content on social media may be associated with self-evaluation and appearance-related intentions. Prior research suggests that exposure to idealized appearance content may provide opportunities for upward appearance comparisons, in which individuals evaluate themselves against perceived superior standards. However, the present study did not measure platform-specific exposure, algorithmic curation, or appearance-related content. Accordingly, upward social comparison is treated here as a theoretical explanatory perspective rather than a directly tested pathway. Such comparisons may be associated with perceived discrepancies between one's actual appearance and socially idealized standards and may relate to beauty ideal internalization and self-objectification, but these processes are not directly identified by the present measure. In recent years, this theory has been extensively applied across gender and age groups to analyze how media-driven aesthetic environments shape individuals’ body attitudes and cosmetic surgery consideration (Jackson and Chen, 2022; Moradi and Huang, 2008).

Within the framework of Objectification Theory, social media use is treated as a contextual factor that may be associated with cosmetic surgery consideration through beauty ideal internalization and self-objectification. The theory provides a basis for examining the proposed relationships among these constructs, but the cross-sectional design does not establish temporal or causal ordering. Accordingly, the present study examines whether the observed associations are consistent with a theoretically specified indirect and sequential pattern. The findings may inform future research on media-related appearance processes, but they should not be interpreted as evidence that the study identified causal psychological mechanisms or intervention effects.

Although prior research has examined indirect psychosocial pathways in the association between social media exposure and behavioral outcomes, many existing models have primarily emphasized distress-related mechanisms or broader self-regulatory constructs, often tested within parallel or serial mediation frameworks (e.g., Kantar et al., 2025; Kök and Kantar 2024). However, comparatively fewer studies have explicitly conceptualized and empirically tested a theoretically ordered cognitive progression from beauty-ideal internalization to self-objectification within a unified objectification-theory framework in explaining cosmetic surgery consideration. By integrating appearance-related social comparison perspectives with objectification-based accounts, the present study extends existing research by examining whether a theoretically specified serial mediation pattern is statistically consistent with the observed associations between social media use and cosmetic surgery consideration. This formulation clarifies the proposed ordering while recognizing that the cross-sectional design cannot establish temporal precedence or causal transmission.

1.1 Social media use and cosmetic surgery consideration

In the present study, social media use refers to general self-reported social media use intensity. This construct is distinguished from appearance-focused social media use, selfie or photo-editing behavior, and exposure to beauty-related content examined in some previous studies. In the present study, cosmetic surgery consideration refers to respondents’ prospective willingness to contemplate cosmetic surgery, rather than general cosmetic-procedure acceptance or completed cosmetic-surgery behavior (Henderson-King et al., 2005). In prior research, passive and active use of Highly Visual Social Media (HVSM) has been associated with cosmetic-procedure consideration, acceptance, and perceived normalization among young people.

Based on a Dutch youth sample, Hermans et al. demonstrated that both passive and active use of HVSM was associated with higher cosmetic surgery consideration, acceptance, and perceived normalization, with appearance-focused activities such as filtering/photo editing exhibiting stronger associations (Hermans et al., 2022). Experimental and correlational studies provide mechanistic insights: Walker et al. (2021) exposed participants to facially cosmetically enhanced images, revealing that while experimental effect sizes were modest, social media use intensity significantly correlated with desire for cosmetic surgery, with body dissatisfaction partially mediating the “social media use → cosmetic surgery desire” relationship. Path analyses focusing on Instagram indicate that image-based activities (e.g., browsing or posting self, celebrity, and peer photos) were associated with cosmetic-procedure acceptance and were indirectly related to acceptance through appearance comparison and body dissatisfaction, with celebrity- and self-focused activities exhibiting particularly strong associations (Di Gesto et al., 2022). Clinical research involving outpatient populations during the pandemic reported that longer social media use and greater engagement with photo-editing apps were associated with a stronger desire to undergo cosmetic procedures (Khan, De La Garza, Lazar, Kennedy, and Vashi, 2024). Systematic reviews further substantiate connections between platform features (visual aesthetics emphasis, selfies) and elevated body dissatisfaction, social appearance anxiety, and cosmetic surgery consideration—affecting both genders (Mironica et al., 2024). Regarding specific behaviors, photo manipulation significantly correlates with facial dissatisfaction and cosmetic attitudes/intentions, predicts the latter two, and associates with appearance-based self-discrepancy (Beos et al., 2021). Responses to appearance-related and influencer-mediated content may vary across audiences and contexts: self-compassion may attenuate the association between social networking-site use and cosmetic attitudes, while adolescents have reported both negative body-image attitudes and body acceptance after viewing thin and curvy models, and Generation Z consumers may critically evaluate influencer content in relation to credibility and trusted sources (Chaihanchanchai and Anantachart, 2026; Conboy and Mingoia, 2024; Ferdousi et al., 2026). Survey evidence from Saudi Arabia has reported associations between social media practices, body dysmorphic disorder (BDD), and acceptance of cosmetic surgery. In an online cross-sectional survey using convenience sampling, Ateq et al. (2024) found that participants with BDD reported higher acceptance of cosmetic surgery than those without BDD. These findings provide contextual support for examining the association between social media use and cosmetic surgery consideration among university students. Therefore, this study proposes the following hypothesis:

H1: Social media use is positively associated with cosmetic surgery consideration.

1.2 The mediating role of beauty ideal internalization

Within social media platforms, users not only passively receive appearance-related information but also gradually internalize “ideal appearance” standards as personal evaluation criteria through ongoing visual consumption. Beauty ideal internalization refers to the process through which individuals adopt socially and culturally promoted standards of physical appearance as personal evaluative criteria or appearance-related goals (Thompson and Stice, 2001). In the present study, this construct is treated as a general appearance-ideal dimension rather than as separate thin-ideal, body-shape, or facial-feature dimensions.

Empirical evidence in social media contexts has reported associations between social media use intensity, beauty ideal internalization, and body-image or appearance-related outcomes, including patterns involving appearance comparison. A large-scale study of young American women demonstrates that social media use intensity significantly predicts body esteem/body satisfaction via chain mediation through beauty ideal internalization and appearance-based social comparison, aligning with the applicability of the Tripartite Influence Model in social media contexts (Jung et al., 2022). Consistently, Instagram-focused research reveals that image-based activities correlate significantly with both appearance comparison and beauty ideal internalization, suggesting that the platform's “seeing and being seen” structure pushes external aesthetic standards into personal self-evaluation systems (Hendrickse et al., 2017). Furthermore, longitudinal and correlational evidence involving HVSM usage associates HVSM engagement with body image concerns and internalization indicators among adolescent/young adult samples, providing contextual support for the overarching “social media → internalization” pathway (Marengo et al., 2018).

More pertinent to this study is that beauty ideal internalization not only connects social media use with body dissatisfaction but also links to more concrete appearance-altering intentions. Research examining beauty-related social media engagement has reported associations with attitudes toward and consideration of cosmetic surgery, with upward appearance comparison and beauty-ideal internalization identified as potential mediating pathways (Seekis et al., 2022). Recent longitudinal evidence further suggests bidirectional associations and time-varying linkages among exposure to beauty ideals on social media, thin-ideal internalization, social appearance anxiety, and cosmetic surgery consideration, highlighting the potential role of internalization in the association between social media use and cosmetic surgery consideration (Yao et al., 2024). Beyond direct exposure-internalization links, socially interactive content likewise drives internalization. Studies examining body talk on social network sites (SNS body talk) indicate that beauty ideal internalization and appearance comparison jointly mediate the “body talk → body dissatisfaction” relationship, uncovering how socially interactive exposure reshapes individuals’ body evaluation benchmarks through internalization (Wang, Geng, Di, Chu, and Lei, 2022). At a broader modeling level, internalization has been robustly associated with cosmetic surgery attitudes: Structural equation modeling grounded in the Tripartite Influence Model confirms that perceived appearance pressure influences cosmetic surgery attitudes via internalization, with significant gender differences in pathway strength (Menzel et al., 2011). Collectively, existing evidence converges to position beauty ideal internalization as a critical psychological conduit through which social media relates to cosmetic surgery consideration. Therefore, this study proposes the following hypothesis:

H2: Social media use is associated with cosmetic surgery consideration through the mediation of beauty ideal internalization.

1.3 The mediating role of self-objectification

In highly visual social media environments, prior theory has linked an observer-oriented perspective on the body to self-objectification (Fredrickson and Roberts, 1997). The “visibility/evaluability” features of social media have been theorized to be associated with viewing the body as an object of external scrutiny, including body surveillance and body shame, which are commonly examined as expressions of self-objectification. Experimental research has reported changes in body surveillance and body shame following a short-term social media abstinence period: adolescent dancers showed lower levels of these outcomes after a 3-day social media fast (Roberts et al., 2022). This finding provides relevant contextual evidence, but it does not establish a causal pathway in the present cross-sectional study.

At the level of daily usage, appearance-focused activities such as taking/posting selfies and pursuing likes are significantly associated with self-objectification. Research on young women indicates that the frequency of posting self-images with objectifying characteristics is jointly predicted by self-objectification and positive feedback (likes), which in turn reinforces an appearance-centric self-presentation cycle (Bell et al., 2018). Adolescent samples similarly reveal that self-objectifying social media use positively correlates with body surveillance and body shame, suggesting this “being looked at” internalizes the observer's perspective as a benchmark for self-evaluation (Salomon and Brown, 2019). Male users are not exempt. Studies with male samples demonstrate systematic associations between Instagram practices and bodily objectification traits/body satisfaction, indicating cross-gender manifestations of social media objectification (Boursier and Gioia, 2022). Crucially, stable links exist between self-objectification and cosmetic surgery attitudes/intentions. Research framed by objectification theory shows variables like body surveillance and body shame predict both “future consideration of cosmetic procedures” and increased “socially motivated reasons” (Calogero et al., 2010). Recent work further identifies self-objectification as influencing alienation and body image inflexibility, leading to appearance-altering intentions through significant chain mediation—revealing the psychological pathway from the viewed self to body modification motives (Huang, Lin, and Yan, 2024). Taken together, prior studies provide a theoretical basis for examining whether general self-reported social media use is associated with cosmetic surgery consideration through self-objectification; the present cross-sectional data do not establish such a causal process. Therefore, this study proposes the following hypothesis:

H3: The indirect association between social media use and cosmetic surgery consideration through self-objectification is positive.

1.4 The theoretically specified serial mediation pattern of beauty ideal internalization and self-objectification

Social media use is associated with beauty ideal internalization. Research on Instagram and highly visual platforms has reported associations between image-centric engagement, beauty ideal internalization, and appearance comparisons, suggesting that the “seeing–being seen” structure of these platforms may be relevant to the incorporation of external standards into self-evaluation systems (Hendrickse et al., 2017; Jung et al., 2022, p. 219). Longitudinal evidence further reveals temporal coupling and bidirectional effects between exposure to social media ideals and thin-ideal internalization, social appearance anxiety, and cosmetic surgery consideration, highlighting the pivotal status of internalization (Yao et al., 2024, p. 22). Regarding internalization's impact on objectification, adolescent longitudinal studies show that sexualized media exposure first increases internalization of appearance ideals, subsequently elevating self-objectification levels at later time points (Vandenbosch and Eggermont, 2016). Evidence from social interaction, self-presentation, and research involving Chinese university students further links appearance-related processes—including body surveillance, body shame, beauty ideal internalization, and appearance comparison—to cosmetic surgery consideration or related body-image outcomes (Galeotti, Marino, Canale, and Vieno, 2024; Wang, Fardouly, Vartanian, & Lei, 2022; Wang, Geng, Di, Chu, and Lei, 2022). Thus, the observed pattern is statistically consistent with indirect associations involving beauty ideal internalization and self-objectification, but it does not establish that social media use influences cosmetic surgery consideration through a causal chain. Within the framework of objectification theory, this sequential ordering is theoretically defensible. Beauty ideal internalization reflects the cognitive incorporation of externally prescribed appearance standards, whereas self-objectification represents a subsequent perceptual shift toward evaluating one's body from an observer's perspective. Accordingly, internalization is treated as the theoretically specified antecedent in the present model, but the cross-sectional design cannot establish that it temporally precedes self-objectification. Prior longitudinal findings are compatible with this ordering, but they do not establish temporal precedence in the present study (Vandenbosch and Eggermont, 2016). Alternative reverse serial and parallel mediation models are also conceptually plausible, but they were not estimated in the present study. The proposed ordering is therefore theoretically specified rather than temporally established, and the present study does not claim that it is empirically superior to these alternatives. Therefore, this study proposes the following hypothesis:

H4: The serial indirect association between social media use and cosmetic surgery consideration through beauty ideal internalization and self-objectification is positive.

In summary, prior research suggests associations among social media use, beauty ideal internalization, self-objectification, and cosmetic surgery consideration. Although existing studies have examined individual and serial indirect associations among these constructs, fewer studies have examined them jointly in a theoretically specified model among Chinese university students. The present study therefore examines whether the observed associations are statistically consistent with a serial indirect pattern involving beauty ideal internalization and self-objectification. Accordingly, the study tests the theoretically specified serial mediation model shown in Figure 1, while recognizing that the cross-sectional design cannot establish causal transmission or temporal precedence.

Although previous studies have examined the relationships among social media use, beauty ideal internalization, self-objectification, and cosmetic surgery consideration, many have focused primarily on female samples. The present study includes male and female university students, but this gender composition is treated only as a sample characteristic and not as an empirical test of gender differences.

Accordingly, the present study examines the associations among social media use, beauty ideal internalization, self-objectification, and cosmetic surgery consideration among Chinese university students. Because no measurement-invariance or multigroup analyses were conducted, the present study does not evaluate gender-specific boundary conditions or cross-gender equivalence of the proposed model.

2 Research methods

2.1 Data sources

Through convenience sampling, this study encompassed over 10 higher education institutions across diverse academic tiers and geographic regions in China, administering questionnaires to university student populations. During data collection, the research team prioritized participants’ privacy protection and right to informed consent. All questionnaires were designed anonymously, excluding personally identifiable information such as names, ID numbers, or contact details. A written informed consent statement at the questionnaire outset explicitly articulated the research purpose, data usage protocols, anonymization procedures, and voluntary participation principle. Data were collected through Wenjuanxing, China's leading survey platform. From 1,200 returned questionnaires, 77 invalid responses were excluded (exclusion criteria: completion time <3 min, inconsistent response patterns, skipped/missing items, etc.). Questionnaires with skipped or missing focal items were excluded during screening; the final analytic sample therefore consisted of the 1,123 retained complete responses, yielding a valid response rate of 93.58%.

The demographic characteristics of the sample are as follows: Regarding gender, 563 participants identified as male (50.1%) and 560 as female (49.9%). In terms of academic level, 752 were undergraduate students (67.0%), 322 were master's students (28.7%), and 49 were doctoral students (4.4%). These descriptive characteristics summarize the composition of the convenience sample and should not be interpreted as evidence of population representativeness.

2.2 Research instruments

This study utilized four psychometrically validated scales to measure core constructs. The evaluation of scale reliability, validity, and model fit referenced established statistical criteria: Cronbach's α coefficient assessed internal consistency; the Kaiser-Meyer-Olkin (KMO) measure determined factor analysis appropriateness; while model error was quantified via Root Mean Square Error of Approximation (RMSEA <0.08 indicating acceptable fit). Overall model fit was evaluated through Comparative Fit Index (CFI), Goodness-of-Fit Index (GFI), and Normed Fit Index (NFI), with values exceeding 0.8 suggesting adequate fit. The χ²/degrees of freedom (χ²/df) ratio maintained a benchmark threshold <3, though its interpretation required conjunctive consideration with RMSEA/CFI/GFI/NFI, particularly in large samples, given the context-dependent variability of these statistical thresholds across research settings and sample characteristics. The full questionnaire items, response options are provided in Appendix A.

2.2.1 Social media use intensity scale

This scale was adapted from the Facebook Intensity Scale developed by Ellison et al. (2007), with selective modifications and reductions to align with the specific research objectives, resulting in a concise 6-item instrument. In the present study, the scale captures general self-reported social media use intensity. It does not specifically assess exposure to appearance-focused content, selfie or photo-editing behavior, or engagement with beauty-related content. It measures social media use intensity through two core dimensions: behavioral integration and emotional involvement, employing a 5-point Likert scale where 1 indicates “Strongly Disagree” and 5 denotes “Strongly Agree”, with higher scores reflecting stronger endorsement. Psychometric analysis confirmed the scale's adequacy: the KMO value of 0.694 demonstrated acceptable sampling suitability for factor analysis, while a Cronbach's α coefficient of 0.805 indicated robust internal consistency and moderate validity. The “social media use” variable was ultimately derived as the mean composite score of all items.

2.2.2 Cosmetic surgery consideration scale

This scale was adapted from the Chinese version of the Acceptance of Cosmetic Surgery Scale (ACSS) translated by Wu et al. (2020) through strategic item reduction. The finalized 6-item subscale eliminates redundant items measuring varying degrees of cosmetic surgery consideration, specifically designed to assess individuals’ prospective willingness to contemplate aesthetic procedures. Utilizing a 7-point Likert scale (1 = “Strongly Disagree” to 7 = “Strongly Agree”), higher scores reflect stronger endorsement with psychometric properties confirming excellent measurement adequacy (KMO = 0.902, exceeding Kaiser's threshold) and exceptional internal consistency (Cronbach's α = 0.956). The “cosmetic surgery consideration” variable was ultimately derived as the mean composite score of all items.

2.2.3 Beauty ideal internalization scale

The scale was adapted from Thompson’s (1999) Sociocultural Attitudes Toward Appearance Questionnaire (SATAQ) through strategic refinement, involving streamlining to 6 core items and replacing original references to “television and magazines” with contemporary “social media” platforms. The adapted items assess respondents’ internalization of socially promoted appearance standards encountered through social media. In the present study, the scale was treated as a general appearance-ideal internalization measure rather than as separate thin-ideal, body-shape, or facial-feature subscales. Measurement employs a 5-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree), where higher scores indicate stronger endorsement. Psychometric validation demonstrated adequate measurement properties with Kaiser-Meyer-Olkin (KMO) sampling adequacy at 0.800 and excellent internal consistency (Cronbach's α = 0.864). The construct of beauty ideal internalization was ultimately derived as the mean composite score of all items.

2.2.4 Self-objectification scale

This study involved the translation and adaptation of two subscales from Lindberg et al.’s (2006) Objectified Body Consciousness Scale for Youth (OBC-Youth): the Surveillance subscale (measuring habitual body monitoring) and the Body Shame subscale (assessing distress about failing to meet beauty standards). The finalized version comprises 9 items evaluating the extent to which participants perceive their bodies as objects of external observation and judgment. Consistent with prior empirical research, including studies on passive social media use and negative body image that operationalized self-objectification as a composite construct, the present study combined the two dimensions for subsequent analyses (Liu, 2023). Responses were recorded on a 7-point Likert scale (1 = “Strongly Disagree” to 7 = “Strongly Agree”), with higher scores indicating stronger endorsement. Psychometric analyses demonstrated excellent suitability for factor analysis (KMO = 0.921) and exceptional internal consistency (Cronbach's α = 0.940). The composite variable “self-objectification” was ultimately derived as the mean composite score of all items. The two subscales were combined because the present study operationalized self-objectification as a composite of habitual body monitoring and appearance-related shame. This operational decision does not imply that body surveillance and body shame are empirically indistinguishable. Because the OBC-Youth was developed for younger populations and the present sample included undergraduate, master's, and doctoral students, the applicability of this youth-derived measure across the university sample should be interpreted cautiously.

2.3 Data processing

Data analysis was conducted using SPSS 26 and AMOS 26. AMOS 26 was used for confirmatory factor analysis and the reported structural model-fit assessment, whereas SPSS 26 was used for correlation analyses. The serial mediation effects were tested separately using PROCESS Model 6 developed by Hayes (2017), based on the raw mean composite scores of the four focal constructs. The bias-corrected bootstrap method was employed to examine the significance of mediation effects, with effects considered statistically significant if the 95% confidence interval did not include zero (Zhou, H., and Long, L. R., 2004). AMOS fit indices and PROCESS regression estimates therefore came from distinct analytic procedures and were interpreted separately.

3 Research results

3.1 Common method bias test

When collecting data via self-report methods, common method bias may occur. To minimize the effects of this bias, the study implemented procedural controls including emphasizing anonymity and confidentiality. Statistically, Harman's single-factor test was employed to examine common method bias in the data (Podsakoff et al., 2003). The results indicated four factors with eigenvalues greater than 1, with the first common factor explaining 42.616% of the total variance—below the 50% critical threshold. Accordingly, Harman's single-factor test was treated only as a limited diagnostic; it does not establish the absence of common method bias, especially because all focal variables were measured by same-source, same-time self-report.

3.2 Confirmatory analyses and correlation analyses

First, confirmatory factor analysis (CFA) was performed using AMOS 26 to assess the reliability and validity of key variables. The Cronbach's α coefficients and composite reliability (CR) values of all major variables exceeded 0.70, with average variance extracted (AVE) values surpassing 0.50 (Table 1), collectively indicating sound reliability. Subsequently, Pearson correlation analyses were conducted between primary variables. Discriminant validity was verified by comparing the square roots of AVE values with inter-variable correlation coefficients. Results demonstrated that the square roots of AVE for all key variables exceeded their respective correlation coefficients with other variables, confirming robust discriminant validity. Thus, the measurement instruments employed in this study meet acceptable standards of reliability and validity.

Table 1. CFA values and correlation coefficients for each variable.

Values along the diagonal represent the square roots of the average variance extracted (AVE). N = 1,123; correlations are Pearson's r.

*denotes p < 0.05, **denotes p < 0.01, ***denotes p < 0.001 (hereafter the same).

Correlational analyses of the key variables revealed significant positive relationships among social media use, cosmetic surgery consideration, beauty ideal internalization, and self-objectification (Table 1). Specifically, social media use demonstrated significant positive correlations with cosmetic surgery consideration (r = 0.102, p < 0.01), beauty ideal internalization (r = 0.492, p < 0.01), and self-objectification (r = 0.196, p < 0.01). Cosmetic surgery consideration was positively correlated with both beauty ideal internalization (r = 0.451, p < 0.01) and self-objectification (r = 0.536, p < 0.01), while a strong correlation emerged between beauty ideal internalization and self-objectification (r = 0.611, p < 0.01). These intervariable correlations provided the empirical foundation for subsequent mediation analyses. To mitigate potential multicollinearity arising from high predictor intercorrelations, PROCESS Model 6 was estimated using the raw mean composite scores; standardized coefficients were reported separately where available. Collinearity diagnostics were also examined. Results indicated variance inflation factors (VIFs) of 1.349, 1.631, and 2.068—all below the critical threshold of 5—confirming the absence of severe multicollinearity (Wen et al., 2018). The data were therefore deemed appropriate for further mediation examination.

3.3 Hypothesis testing

Chain mediation analysis examines whether an independent variable is statistically associated with a dependent variable through multiple mediators specified in sequence. This procedure typically employs multiple regression analysis to ascertain correlations among mediators and the outcome variable, with indirect effects calculated to verify the chain mediation pathways. The structural model yielded χ²/df = 7.77, p < 0.001, RMSEA = 0.078, GFI = 0.872, TLI = 0.917, and CFI = 0.930. These indices refer to the AMOS structural model and are separate from the composite-score PROCESS analysis. Although the overall fit evidence was mixed, several key indices indicated an acceptable or marginally acceptable level of fit. The overall fit evidence was mixed, and these indices were therefore interpreted cautiously in evaluating the hypothesized structural associations.

Subsequently, the PROCESS macro developed by Hayes was employed to evaluate the chain mediation effects of beauty ideal internalization and self-objectification between social media use and cosmetic surgery consideration. Using 5,000 bootstrap resamples, 95% confidence intervals (CIs) were computed for the mediation effects. Statistical significance was determined based on whether the 95% CI included zero; effects were deemed significant if the interval excluded zero (Fang and Zhang, 2012). Results of the chain mediation analysis are presented in Table 2.

Table 2. Regression equations for chain mediation analysis of the model.

N = 1,123. B = unstandardized coefficient; β = standardized coefficient. PROCESS Model 6 was estimated using raw mean composite scores. Model R² values were 0.2417 for Beauty-Ideal Internalization, 0.3870 for Self-Objectification, and 0.3192 for Cosmetic Surgery Consideration.

As illustrated in Table 3, the results showed that social media use had a significant positive total association with cosmetic surgery consideration (B = 0.1918, SE = 0.0557, p = 0.0006), supporting Hypothesis H1. The 95% confidence interval (CI) of the total effect was [0.0825, 0.3011], excluding zero, indicating a statistically significant but small overall association. After introducing the mediators (beauty ideal internalization and self-objectification), the direct effect of social media use remained significant (B = −0.1926, SE = 0.0537, p = 0.0004, 95% CI [−0.2981, −0.0872]). This sign reversal, together with oppositely directed specific indirect effects, is more appropriately described as inconsistent mediation and may involve statistical suppression rather than ordinary partial mediation. Notably, some specific indirect effects were opposite in sign to the direct effect, indicating an inconsistent mediation pattern that may involve statistical suppression (Wen and Ye, 2014). When indirect and direct effects have opposite signs, their influences cancel each other out, causing the total effect to be obscured. Accordingly, this pattern should be interpreted cautiously as suppression-like rather than as evidence of a single mediating mechanism. The chain mediation operated through three primary pathways:

• Ind1: Social media use → Beauty ideal internalization → Cosmetic surgery consideration (95% CI = [0.1587, 0.3216], excluding zero; indirect effect = 0.2378), supporting Hypothesis H2;
• Ind2: Social media use → Self-objectification → Cosmetic surgery consideration (95% CI = [−0.1450, −0.0626], excluding zero; indirect effect = −0.1024), which was opposite to the hypothesized positive direction; therefore, H3 was not supported as hypothesized;
• Ind3: Social media use → Beauty ideal internalization → Self-objectification → Cosmetic surgery consideration (95% CI = [0.1916, 0.3118], excluding zero; indirect effect = 0.2491), supporting Hypothesis H4.

Table 3. Bootstrap test for mediation effects.

N = 1,123. Estimates are unstandardized. For indirect effects, SE is the bootstrap standard error and inference is based on 5,000-sample bias-corrected bootstrap 95% confidence intervals, PROCESS does not report a conventional p value for these indirect effects.

Note:

Ind1: Social media use →Beauty ideal internalization→ Cosmetic surgery consideration;

Ind2: Social media use → Self-objectification → Cosmetic surgery consideration;

Ind3: Social media use →Beauty ideal internalization→ Self-objectification → Cosmetic surgery consideration.

4 Discussion

4.1 Social media use and cosmetic surgery consideration

This study found a statistically significant positive total association between general self-reported social media use intensity and cosmetic surgery consideration. The zero-order correlation was small (r = 0.102), and the simple model explained only 1.05% of the variance (R² = 0.0105). Therefore, statistical significance should not be interpreted as a large practical effect, and the result should not be understood as evidence that general social media use directly increases cosmetic surgery consideration. The platform-specific features described below are drawn from prior research and are not measured by the general self-reported social media use intensity measure used in the present study. From a media-environment perspective, contemporary social platforms are highly visual, appearance-oriented, and algorithmically curated. Platforms such as Instagram, TikTok, and Snapchat prioritize image-centered content, short-form videos, and beautified self-presentations. Their technological affordances—including filters, photo-editing tools, and personalized recommendation systems—may increase exposure to curated and appearance-focused content. Within such visually saturated environments, physical appearance becomes a salient dimension of online interaction and self-presentation. Furthermore, social media may create more opportunities for appearance-based social comparison. Unlike traditional mass media, social platforms are embedded within users’ peer networks, making comparison targets socially proximal and psychologically relevant. Frequent exposure to idealized representations of peers, influencers, and celebrities may heighten awareness of discrepancies between one's actual and desired appearance. In such contexts, cosmetic surgery may be cognitively framed as a socially visible and practically attainable strategy for appearance enhancement. Existing studies have reported related associations in platform-specific or appearance-focused contexts. Dane and Bhatia (2023) reported associations between social media engagement and appearance-related distress across cultural contexts. Longitudinal research by Yao, Niu, and Sun (2024) further demonstrated that exposure to idealized imagery on social media significantly predicted subsequent cosmetic surgery consideration. These prior findings provide context for the observed total association, although the effect in the present study was small. However, it is important to note that social media environments are not unidirectional in their influence. Although prior literature has discussed both appearance-focused and corrective content, these forms of exposure were not measured in the present study and cannot explain the observed negative direct association. The negative direct association should therefore be interpreted as an unresolved statistical pattern that may reflect inconsistent mediation, statistical suppression, omitted variables, or other uncontrolled influences. These possible explanations warrant further examination in future longitudinal or experimental research.

4.2 The mediating role of beauty ideal internalization

The findings showed that social media use was positively associated with beauty ideal internalization, which was in turn positively associated with cosmetic surgery consideration; the corresponding specific indirect association was positive and statistically significant. This pattern is consistent with beauty ideal internalization as a statistical mediator in the observed relationship between social media use and cosmetic surgery consideration. Objectification Theory (Fredrickson and Roberts, 1997) provides a robust theoretical framework for interpreting this pathway. The theory proposes that repeated exposure to appearance-focused representations in sociocultural contexts encourages individuals to adopt externally defined beauty standards as self-evaluation criteria. When such standards are internalized, individuals begin to assess their own bodies through an observer's perspective, increasing appearance monitoring and evaluative concern. The platform-specific characteristics discussed below are drawn from prior literature and were not measured by the general self-reported social media use intensity measure used in the present study. In contemporary social media environments, this internalization process may be intensified by several structural characteristics. First, platforms such as Instagram, TikTok, and Snapchat are highly visual and algorithm-driven, may amplify content that conforms to dominant aesthetic norms. Second, interactive mechanisms—including likes, comments, and shares—publicly quantify appearance-based approval, which may reinforce perceptions of socially endorsed beauty standards. Third, users actively participate in producing and disseminating curated images, potentially contributing to the reproduction of beauty norms as collectively circulated social scripts. Through repeated exposure, social endorsement, and participatory reproduction, idealized appearance standards become cognitively embedded as personal benchmarks. Once internalized, discrepancies between perceived appearance and internalized standards may be associated with body dissatisfaction and cosmetic surgery consideration. In this context, cosmetic surgery consideration may be associated with perceived discrepancies between internalized ideals and actual appearance; however, the present study did not assess completed cosmetic-procedure behavior or the motivations underlying such consideration. Recent empirical research has reported related associations. Longitudinal evidence (Yao, Niu, and Sun, 2024) reported an indirect association between exposure to idealized social media content and subsequent cosmetic surgery consideration through beauty ideal internalization. Similarly, O’Connor et al. (2024) found that appearance-related platform features exacerbate perfectionistic standards and upward comparison, further reinforcing internalization processes. Together, these findings provide contextual support for examining beauty ideal internalization as a statistical mediator in the association between social media use and cosmetic surgery consideration, but they do not establish a causal mechanism in the present study.

4.3 The mediating role of self-objectification

The findings indicated a statistically significant but negative specific indirect association through self-objectification alone between general self-reported social media use intensity and cosmetic surgery consideration, opposite to the hypothesized positive direction. Although the concept of “mirror gaze” offers a relevant interpretive framework for understanding how external perspectives may shape body surveillance and appearance-related intentions, the present cross-sectional data do not establish such a psychological process. Prior studies have reported related positive associations in more specific media-use or appearance-focused contexts. For example, Çınaroğlu and Yılmazer (2025) reported that problematic social media use was positively associated with self-objectification, which was in turn associated with body-image disturbance in a Turkish sample. Similarly, research involving young Chinese women reported an association between selfie-editing behavior, self-objectification, and cosmetic surgery consideration, with self-objectification examined as a mediator (Sun, Q., 2020). These findings provide contextual support for the relevance of self-objectification in appearance-focused media research, but they do not explain the negative self-objectification-only indirect association observed in the present study.. From a media-psychology perspective, social media may provide opportunities for users to evaluate themselves through perceived external viewpoints. Visual content and social interactions, such as likes and comments, may be associated with self-evaluation through an “other-assessment lens.” However, these platform-specific processes were not measured in the present study and cannot be identified as explanations for the observed indirect association. Accordingly, the present cross-sectional findings cannot establish that self-objectification develops over time or propels cosmetic surgery consideration. Prior literature has conceptualized self-objectification as one pathway through which social media relates to cosmetic intentions. Accordingly, the present result should be interpreted as an inconsistent, suppression-like indirect association.

4.4 The chain mediation effect of beauty ideal internalization and self-objectification

The model results indicated a statistically significant positive association between beauty ideal internalization and self-objectification, and self-objectification was in turn significantly associated with cosmetic surgery consideration. The positive serial indirect effect (effect = 0.2491, 95% CI [0.1916, 0.3118]) is consistent with the hypothesized ordering from beauty-ideal internalization to self-objectification. Theoretically, this chain relationship aligns with the logical sequence proposed by objectification theory: “appearance discipline influences appearance-related behavioral intentions through ideal internalization followed by self-objectification” (Fredrickson and Roberts, 1997). It also resonates with the Extended Sociocultural Model's emphasis on phased psychological mechanisms, wherein sociocultural pressures initially lead to the internalization of ideal standards, which then act upon downstream behaviors through objectification components such as body surveillance and shame (Tiggemann and Slater, 2013). Although alternative plausible pathways—such as reverse ordering (self-objectification preceding internalization) or parallel mediation structures—may also be conceptually defensible, the present ordering is specified based on objectification theory's proposed progression from internalized appearance standards to observer-oriented self-perception. The findings are therefore consistent with the theoretically specified sequential structure within the present framework.

Beyond the positive serial indirect association, the results revealed an inconsistent mediation pattern that may involve statistical suppression: the total association was positive, whereas the direct association became negative after the mediators were included. Because corrective narratives were not measured in this study, the negative direct association cannot be attributed to body positivity campaigns, surgical-risk disclosures, authenticity initiatives, or other buffering exposures. It may instead reflect omitted variables, measurement overlap, statistical suppression, or other uncontrolled factors. These possibilities should be tested directly in future research. Such countervailing statistical associations are consistent with suppression patterns described in mediation analysis (MacKinnon et al., 2000; Wen and Ye, 2014) and align with recent research emphasizing complex serial mediation processes in digital media contexts (Kantar et al., 2025). Similar structured psychosocial mediation patterns have also been documented in adjacent domains where self-related cognitive mechanisms connect contextual influences to behavioral outcomes, further supporting the broader applicability of ordered indirect pathways in contemporary psychological research (Kantar and Yılmazcan, 2025). However, unlike distress-centered serial mediation models that primarily emphasize emotional dysregulation or regulatory deficits, the present study highlights a cognitively structured pattern grounded in objectification theory. Specifically, it distinguishes the cognitive incorporation of beauty ideals from the observer-oriented self-perception reflected in self-objectification, thereby refining existing serial mediation frameworks by clarifying the theoretically specified relationships among beauty ideal internalization, self-objectification, and cosmetic surgery consideration. Accordingly, the present data support an inconsistent statistical pattern but do not establish specific risk-versus-protective mechanisms within social media environments.

4.5 Research implications

Based on objectification theory and the extended sociocultural model, the findings indicate that social media use relates to cosmetic surgery consideration through the chain mediating effects of beauty ideal internalization and self-objectification. Notably, an inconsistent mediation pattern was identified. Because the study did not measure corrective content or other buffering exposures, the source of the negative direct association remains uncertain; such explanations should be treated as hypotheses for future research.

From a health communication perspective, the present findings extend objectification-based psychological research by identifying appearance-related self-evaluative processes as relevant targets for message design. In particular, media-literacy messages may help audiences recognize that beauty standards are socially constructed, evaluate the credibility and context of appearance-related content, and distinguish general social media use from exposure to specific appearance-focused messages. More broadly, meta-analytic evidence suggests that stand-alone interventions can produce small but reliable improvements in body image (Alleva et al., 2015). Because the present study did not measure message exposure, media-literacy training, or intervention outcomes, these implications should be understood as theoretically informed directions for future communication research rather than evidence of intervention effectiveness. The inconsistent statistical pattern further suggests that health messages should not assume that social media has a uniform influence; future counter-messaging should be designed and tested to address potentially competing appearance-related influences within the same media environment.

The findings also have implications for future health communication research. Because the present study did not measure algorithmic recommendation, appearance-focused exposure, corrective narratives, message exposure, or intervention outcomes, it cannot establish that filter disclaimers, algorithmic-transparency measures, changes to appearance-focused recommendations, surgical-risk information, media-literacy education, or institutional programs would reduce beauty-ideal internalization, self-objectification, or cosmetic surgery consideration. These strategies may be examined as testable possibilities in future longitudinal, experimental, and implementation research. Such research should compare different message designs and assess whether they are associated with changes in the relevant psychological outcomes before claims about intervention effectiveness are made.

5 Limitations and future research

This study identified meaningful associations among social media use, beauty-ideal internalization, self-objectification, and cosmetic surgery consideration, while several limitations warrant attention. First, methodologically, the reliance on convenience sampling and cross-sectional design may constrain the generalizability and causal inference capacity of the conclusions. Reverse causality and reciprocal relationships also remain plausible; individuals with greater appearance concern may use social media more intensely, seek appearance-related content, or report stronger self-objectification. Future studies should expand sample diversity across institutional types and demographic groups, employing longitudinal tracking or experimental designs to enhance external validity and elucidate dynamic causal pathways. Second, all focal variables were measured using same-source, same-time self-report measures. Harman's single-factor test is only a limited diagnostic, so common method variance cannot be ruled out. Incorporating behavioral observation and physiological indicators would improve objectivity. Additionally, the general social media use intensity measure does not distinguish appearance-focused from non-appearance-focused use, active from passive use, or platform-specific exposure; future studies should refine these metrics. The cosmetic surgery consideration measure also did not distinguish surgical from nonsurgical procedures, specify a behavioral time frame, or assess participants’ previous cosmetic-procedure experience; future studies should address these dimensions. The translation and adaptation procedures are not described in full in the present report. Although the complete adapted items are provided in Appendix A and the available reliability and validity indices are reported in Table 1, the complete χ² and df values and SRMR for the AMOS structural model, separately documented expert assessment, pilot testing, alternative measurement models, complete item-level factor-loading results, a detailed survey period, and a full age distribution are not reported. These reporting limitations should be addressed in future validation studies. Finally, while the current model identified the mediating roles of beauty ideal internalization and self-objectification, integrating additional variables—such as self-esteem, peer influence, cultural values—and moderators like gender, age, and media literacy would advance a more comprehensive theoretical framework.

6 Conclusion

This study, based on a sample of Chinese university students, clarifies the complex pathways through which social media use relates to cosmetic surgery consideration. First, general self-reported social media use intensity showed a statistically significant positive total association with cosmetic surgery consideration. However, when mediating variables are incorporated, the direct effect becomes negative. This pattern is consistent with countervailing statistical associations. Second, beauty ideal internalization showed a significant positive specific indirect association between social media use and cosmetic surgery consideration. This pattern is consistent with beauty-ideal internalization as a relevant correlate within the association between social media use and cosmetic surgery consideration. Third, the self-objectification-only specific indirect association was significant but negative. Fourth, the serial indirect association through beauty-ideal internalization and self-objectification was positive and statistically significant. This pattern was statistically consistent with the theoretically specified ordering, although the cross-sectional design cannot establish temporal precedence. From a practical perspective, these findings highlight the importance of strengthening critical media literacy and fostering balanced attitudes toward digitally mediated beauty norms. Such efforts may help mitigate appearance-related distress and support healthier, more reflective body-related decisions among Chinese university students.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Jiankun Gong, University of Malaya, Malaysia

Reviewed by: Le Cao, University of Erlangen Nuremberg, Germany

Rong Han, Northwestern Polytechnical University, China